This proposal represents a collaborative effect between Fox Chase Cancer Center, US Healthcare and Mathematic Policy Research (MPR) to reduce mortality from breast cancer in a geographically and organizationally-defined group of 43,000 women 50 years of age and older. The study aims are improve compliance with prescribed medical recommendations; improve availability and quality of state-of-the-art health services, assure continuity of care, from detection through management, and reduce barriers to access. The research builds upon the foundation of US HEALTHCHECK, a program of breast and colorectal screening for members of HMO PA/NJ who are aged 40 to 64. We aim to create an ongoing system for stimulating patient and physician compliance with the recommended behaviors and ensuring the appropriate medical management of patients with suspicious findings. We seek to reduce the patient, physician and health system barriers that result in avoidable mortality. Since the financial barrier to mammorgraphy will be removed for the study participants, we can study the efficacy of interventions needed to maximize physician and patient compliance. The interventions include complementary strategies aimed at women members, their primary physicians and associated radiologist and at the health care system itself. We will use a stepped-care approach, starting initially with the optimal combination of strategies to reach the most women at the least cost. Then, we will add additional, more intensive strategies to reach the more hard-to-reach women. The proposed physician interventions include a self-learning manual, workshops, a comprehensive reminder system that will serve as the bridge between the patient and physician interventions and a peer leader system. The efficacy of the interventions will be evaluated by means of 4 surveys conducted by MPR. In each wave, 300 physicians (150 randomly selected HMO PA/NJ physicians and 150 randomly selected geographical controls) and 1,000 age-eligible women will be interviewed (500 randomly selected HMO PA/NJ and 500 age-eligible geographic controls). Analytic techniques will include logit and probit models and other appropriate techniques. Interventions with proven success will be incorporated into the accepted standard of practice and will not terminate, as so often happens, at the end of the grant. With a nationwide membership of one million US Healthcare members, and a medical practice similar to fee-for-service medicine, the potential of this project to alter the practice of medicine is great.